{"record": {"aliases": [], "artifacts": [{"id": "arxiv:2609.04867", "kind": "paper", "url": "https://arxiv.org/abs/2609.04867"}], "authors": ["Yuchen Sun", "Qian Yang", "Jun Wang", "Detai Xin", "Guoqiao Yu", "Guanglu Wan", "Qi Jia"], "authors_reference": {"basis": "arxiv_metadata", "retrieved_at": "2026-11-03", "source_key": "claire-radar:2609.04867", "source_url": "https://arxiv.org/abs/2609.04867"}, "categories": ["General AI", "Vision & 3D", "cs.MM"], "description": {"en": "Evaluates text-to-audio-video generation with an audio-centric diagnostic benchmark comprising 900 human-verified samples, factorizing audio evaluation by audio type and sound-source visibility, and assessing four perceptual dimensions via 35 fine-grained criteria using an MLLM-as-a-Judge protocol."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://arxiv.org/abs/2609.04867", "source": "claire_radar", "source_url": "https://arxiv.org/abs/2609.04867", "title": "PRISM-Bench"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:2609.04867", "modality": null, "name": "PRISM-Bench", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"confidence": "High", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-indexed", "display_eligible": "false", "origin_record_id": "bm_prism-bench_97126e49", "origin_source": "arxiv", "recognition_confidence": "1.0", "record_sha256": "cac7f4b7a89061ed51004213cced477b3e485aebb9110246219859da12772cc6", "relation": "introduces", "review_model": "deepseek-v4-pro", "review_state": "ai-name-audit-deferred", "reviewed_at": "2026-09-08T00:56:29.478814Z", "source_url": "https://arxiv.org/abs/2609.04867"}, "publisher": null, "released": "2026-09-04", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:2609.04867", "source_url": "https://arxiv.org/abs/2609.04867"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-2609-04867", "source": "claire_radar", "source_benchmark_id": "2609.04867", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": [], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Vision & 3D", "attention": {"asOf": "2026-10-05", "attemptedAt": "2026-10-04T15:31:48.030916Z", "githubRepo": null, "githubScope": null, "githubStars": null, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": "https://huggingface.co/papers/2609.04867", "observedAt": "2026-10-04T03:02:42.754940Z", "signalStatus": {"githubStars": {"state": "not_applicable"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"checkedAt": "2026-10-04T15:31:48.030916Z", "state": "unavailable"}}, "status": "unavailable"}, "attentionForecast": {"confidence": "Medium", "horizon": "7d", "reason": "The benchmark's focus on audio evaluation for generative models is niche but increasingly relevant, likely attracting researchers in audio-visual generation and evaluation.", "score": 58}, "capabilities": [], "confidence": "High", "construction": "Transform Existing", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-08T08:59:06.803818Z", "inputHash": "539fceceae36139a3896138e6bce7b69a75e210a16fe0d4cc3036673c28ffd97", "model": "deepseek-v4-flash"}, "curation": {"decisionReason": "Neither a code repository nor dataset release is mentioned; without public artifacts, the evaluation protocol cannot be independently reproduced.", "model": "deepseek-v4-pro", "policy": "source-evidence-v2", "reviewedAt": "2026-09-08T00:56:29.478814Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-indexed", "demo": false, "description": "Evaluates text-to-audio-video generation with an audio-centric diagnostic benchmark comprising 900 human-verified samples, factorizing audio evaluation by audio type and sound-source visibility, and assessing four perceptual dimensions via 35 fine-grained criteria using an MLLM-as-a-Judge protocol.", "displayEligible": false, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "public_reusable", "evidence": {"reasonCodes": ["coined title prefix ending in Bench or Benchmark", "benchmark release stated in one sentence", "evaluation protocol evidence"], "snippet": "We present PRISM-Bench, the first audio-centric diagnostic benchmark for T2AV generation."}, "familyId": "bmf_ea614005cbac", "firstSeenAt": "2026-09-07", "heat": null, "id": "bm_prism-bench_97126e49", "indexedAt": "2026-09-07T00:39:11.978130Z", "industrySectors": [], "links": {"code": null, "data": null, "paper": "https://arxiv.org/abs/2609.04867", "pdf": "https://arxiv.org/pdf/2609.04867", "project": null, "report": "https://arxiv.org/abs/2609.04867"}, "metrics": [], "motivation": "Text-to-audio-video (T2AV) generation has advanced rapidly, but its evaluation still underestimates the audio modality.", "name": "PRISM-Bench", "oneLine": "Evaluates text-to-audio-video generation with an audio-centric diagnostic benchmark comprising 900 human-verified samples, factorizing audio evaluation by audio type and sound-source visibility, and assessing four perceptual dimensions via 35 fine-grained criteria using an MLLM-as-a-Judge protocol.", "paperTitle": "PRISM-Bench: An Audio-Centric Diagnostic Benchmark for Text-to-Audio-Video Generation", "primaryDomain": "General AI", "publication": {"evidence": "19 pages, 10 figures, 4 tables. Accepted at ACM Multimedia 2026 (MM '26). This arXiv version includes supplementary appendices not included in the conference proceedings version", "evidenceLevel": "author-claim", "evidenceUrl": "https://arxiv.org/abs/2609.04867", "source": "arxiv-comments", "status": "acceptance_claimed", "venue": "ACM Multimedia 2026 (MM '26)", "verifiedAt": "2026-09-07T00:39:11.978130Z"}, "publications": [], "ranking": {"90d": {"confidence": "Low", "coverage": 0.0, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 90}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": null, "value": null}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "rank": null, "score": null}, "method": "attention-ranking-v13", "rank": null, "score": null}}, "readiness": "Paper only", "recognitionConfidence": 1.0, "relation": "introduces", "releasedAt": "2026-09-04", "source": {"authors": ["Yuchen Sun", "Qian Yang", "Jun Wang", "Detai Xin", "Guoqiao Yu", "Guanglu Wan", "Qi Jia"], "categories": ["cs.MM", "cs.AI", "cs.SD"], "id": "2609.04867", "title": "PRISM-Bench: An Audio-Centric Diagnostic Benchmark for Text-to-Audio-Video Generation", "type": "arxiv", "url": "https://arxiv.org/abs/2609.04867"}, "sourceUpdatedAt": "2026-09-04T08:27:10Z", "topics": ["cs.MM"], "venueAttempts": [{"decisionRaw": "19 pages, 10 figures, 4 tables. Accepted at ACM Multimedia 2026 (MM '26). This arXiv version includes supplementary appendices not included in the conference proceedings version", "evidence": [{"level": "author-claim", "observedAt": "2026-09-07T00:39:11.978130Z", "rawValue": "19 pages, 10 figures, 4 tables. Accepted at ACM Multimedia 2026 (MM '26). This arXiv version includes supplementary appendices not included in the conference proceedings version", "sourceType": "arxiv-comments", "sourceUrl": "https://arxiv.org/abs/2609.04867"}], "reviewStatus": "accepted", "venueName": "ACM Multimedia 2026 (MM '26)"}], "whyItMatters": "Fills the evaluation gap for audio quality in T2AV generation, providing a granular diagnostic tool to compare models on audio coherence, quality, expressiveness, and prompt following, guiding improvement in audio-centric aspects."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
